200k?

#5
by MarsupialAI - opened

Is this based on the 200k context version of Yi?

Is this based on the 200k context version of Yi?

No this is the base model. The 200k context version of Yi is repetition problems.

Locutusque changed discussion status to closed

Yes, I'm aware of the repetition issues in the 200k model. I've been hoping that would be mitigated to some extent by finetuning.

Sign up or log in to comment